I posted this suggestion before and it didn’t get much traction, but with the US Postal Service bleeding money and probably eliminating Saturday delivery it’s worth bringing out once more. The basic idea is to make mail more like e-mail so that postal addresses are easier to remember and don’t have to change when people move.
Here’s how DARPA elegantly expressed the basic way to send information on the internet back in 1981:
A distinction is made between names, addresses, and routes. A name indicates what we seek. An address indicates where it is. A route indicates how to get there.
It’s the distinction between names and addresses that’s important. With email, the name is an email account and the address is a number called an IP address that identifies the server hosting the account. So if you wanted to send a message to me, for example, the name would be “email@example.com” and the address would be a series of digits like “123.456.789.” You only need the name, not the address. Databases called the DNS registry link the name to the correct IP address so that the message gets to my inbox.
This is great because it allows people to send me messages even if they have no idea where my server is. They don’t have to know what my hard-to-remember IP address is and I don’t have to inform them if I move the site to a new host. I just need to tell the DNS registry that the IP address has changed and it takes care of the matter for everyone.
Physical mail doesn’t work this way. Unlike email servers, the US Postal Service collapses name and address into one unit. If you want to send me a message via USPS you can’t just write my name on an envelope, you also need to know my current physical address. This is a huge inconvenience! It means that every time I move I have to inform everyone who sends me mail that my address has changed and they have to update their address books. It’s a time-consuming and error-prone process.
Does it have to be this way or is the technology cheap enough now that we could make mail more like email? From my original post:
Why should we have to remember cumbersome physical addresses and update all our contacts when we move? It would be a lot easier to simply use the equivalent of a domain name address and associate it in a database with a physical mailing location. Call it a Postal Name System (PNS). Everyone could have their own, easily memorized address to use for life. When people move, they just notify the PNS of the change and their postal name keeps functioning seamlessly, associating their postal address with their new physical location.
In other words, there’s no longer any reason why the physical locations where we live and work should have anything to do with the postal addresses people use to send us stuff.
One’s email address and postal address could even be the same thing. When your server changes you’d tell the DNS, when your home changes you’d tell the PNS. Regardless of whether correspondents wanted to send you an email or a letter, all they’d need to know is one easily remembered name that never needs to change.
The obvious objection is that postal workers still need a physical address printed on the package to make the final delivery, but USPS already deals with this problem for automated mail forwarding. Machines scan an address, match it against a forwarding database, and put on a corrected label if forwarding has been requested. The idea proposed above could use a similar process on a much larger scale.
A second problem is how we transition from our current addressing system to the new one. This seems solvable by the methods discussed in the post and comments on this forum. International shipping would create difficulties, so the system might have to be limited to locations in the US, at least initially.
A benefit would be a new revenue stream for USPS if the service charges to register names. If it’s offered as a premium service I think many people would gladly pay a reasonable amount for it. They might even register multiple names for the same reasons they have multiple email addresses, separating them as personal or work, public or private. A DNS-like system wouldn’t have to replace the current one, it could just work on top of it for people who opt-in.
Is this idea too good to be true? Possibly. I don’t want to feign expertise in an industry I know little about and the fact that it hasn’t already been implemented suggests there must be other obstacles. I’m curious what those are; perhaps printing new labels is too expensive and time consuming? On the other hand, it might be an idea that’s just now becoming technologically feasible. With the Postal Service in such bad shape it’s worth considering. It won’t make up for the fact that email has largely replaced sending physical messages, but it would reduce the transaction cost of using the mail.
Additional notes: There’s one blog devoted to promoting this idea. Unfortunately it stopped after one post. The Intelligent Mail program offers address correction services to some companies. Intriguingly, USPS once sought control of the .us top-level domain. Might they have been considering a similar evolution more than a decade ago?